Why do we buy what we buy?
Why on earth would anyone buy the infamous leg lamp (from t
he movie, A Christmas Story)? If you don’t know what motivates someone to buy your product, how can you market it effectively? Next time you conduct a survey, plan ahead to use some of these are classic analytic tools. They make it easy to determine what drives purchase interest. They are easy to deploy, and interpret. There’s a deeper treatment of some of these tools, and others, in a nice Quirk’s article by Kevin Gray.
1. Get a foothold on what matters with Key Drivers Analysis: Respondents rate a product on how well it delivers on a battery of product attributes. You could just look at what attributes are rated most highly, but that won’t tell you how much each matters in driving purchase of the product. Correlate those scores with some measure of overall performance, such as purchase intent, and you’ll know what really matters to your customers. When you do this correlation, the score you get for each attribute is called its Derived Importance. The official term: Pearson Product Moment Correlation. Here’s an example on benefits people might seek in a toy (all examples are hypothetical).
2. Find out what drives customers head over heels with Regression: A more sophisticated route is to use multivariate regression. It takes into account all the attributes at once, not one by one (as in correlation), and so can be more powerful. It gets complicated, and can be treacherous, so you’ll want a statistically savvy person to advise you on this. In this example, Regression identifies significant opportunities to increase satisfaction with this store’s shopping environment. In an actual study, this factor, “shopping environment” would be comprised of several attributes, with data on each one, so Client gets clear direction as to what aspects of the shopping environment to fix.
3. Get a leg up on the competition…with Quadrant Analysis: For each attribute, plot 2 numbers on a classic x,y chart: its Derived Importance and the performance (either mean rating, or more commonly, the % giving it a high rating). You can easily see how your product is performing on the characteristics that matter to people. Do this simple analysis for your product, and key competitors, and you can see your strengths and weaknesses relative to them. In these Opportunity Maps, every attribute lands in one of these 4 quadrants, based on the Derived Importance, and Performance score.
4. Avoid the Achilles Heel of Stated Importance. People rate a battery twice, once on the importance and once on performance, but there are at least 3 downsides. That’s why we typically use Derived Importance, not Stated Importance. These are the downsides of Stated Importance:
- “everything is important” – you may get little discrimination in scores
- people say what is socially correct – a correlation sometimes is far more revealing
- it is tedious for the respondent to rate a battery of attributes once on importance, and then again on performance. That erodes data quality.
5. Shake a leg, and find Opportunities with Need Gap Analysis: There is at least one situation where you want to ask people to rate attributes on importance (Stated Importance). Sometimes we are looking to find holes in the market, attributes that people want and can’t get. Need Gap Analysis enables us to do that. This is an approach that lets you drill down to measure performance on attributes among the people who really care about that attribute. We find out how many people really care about Attribute X and are disappointed with the delivery of current products on that attribute.
How to: Everyone rates the battery on importance (Stated Importance). Everyone rates the battery again on performance (usually performance of the brand they use most often). Now, for each attribute, filter the data down to those who rated this attribute highly important. Find out how this subset of people rated the performance of the brand they use most. The Need Gap is defined as 100 – % Top 2 Box performance (among those who rated this important). Plot: X = the size of the Need Gap. Y = the percentage of respondents who rated this important. Attributes in the upper right quadrant are ones with large Need Gaps: a lot of people care about this attribute, and few people are very happy with their existing products. These disgruntled people are showing you your opportunity area.
